Mobile edge computing (MEC) can alleviate the computing resource shortage problem of mobile user equipment (UEs). However, due to long communication distance or the obstruction of big obstacles, the direct communication link may not exist between a UE and a MEC node. It thus hinders the task offloading in MEC. Unmanned aerial vehicles (UAVs) have high degree of mobility and can carry lightweight computation and storage modules. This paper presents a UAV-assisted MEC method, in which the UAV can relay the task-input data of a UE to the MEC node and can also utilize the airborne computation and storage resource to shorten the execution time of the offloaded tasks. Considering the strict order dependency among multiple offloaded tasks, this paper optimizes the task scheduling and the UAV flight path in a joint manner. A heuristic algorithm based on particle swarm optimization (PSO) is also developed to find the optimal solution. The simulation results show that the proposed multitask scheduling method can always find the best tradeoff between the UAV’s position and the wireless channel condition. In comparison to the other three baseline scheduling methods, the proposed method can use the minimum execution time to complete all the offloaded tasks.

中文翻译：

---

无人机辅助移动边缘计算中用于完成时间最小化的高效多任务调度

移动边缘计算（MEC）可以缓解移动用户设备（UE）的计算资源短缺问题。然而，由于长距离通信或大障碍物的阻碍，在UE和MEC节点之间可能不存在直接通信链路。因此，这阻碍了MEC中任务的卸载。无人机（UAV）具有高度的机动性，可以携带轻量级的计算和存储模块。本文提出了一种无人机辅助的MEC方法，该方法可以将UE的任务输入数据中继到MEC节点，还可以利用机载计算和存储资源来缩短卸载任务的执行时间。考虑到多个卸载任务之间的严格顺序依赖性，本文以联合方式优化了任务调度和无人机飞行路径。还开发了一种基于粒子群优化（PSO）的启发式算法，以找到最佳解决方案。仿真结果表明，所提出的多任务调度方法总能在无人机位置和无线信道状况之间找到最佳折衷。与其他三种基线调度方法相比，所提出的方法可以使用最少的执行时间来完成所有卸载的任务。